Adoptive cell therapy of autologous tumor-infiltrating lymphocytes (TIL) can now mediate objective cancer regression in 50% of patients with metastatic melanoma. While TIL therapy has made incremental improvements over recent decades, the fundamental methodologies have not significantly changed despite major advances in the field of immuno-oncology. Current protocols for TIL therapy focus on the ex vivo expansion of TIL from resected tumor biopsies using high-dose interleukin-2 over many weeks. This is a time- consuming and expensive process that limits the eligibility of many patients and reduces the number of cancer treatment facilities capable of offering this life-saving therapy. The most important factors governing clinical response are the anti-tumor reactivity of the final TIL infusion product, and the total number of TIL generated for adoptive transfer. Current methodologies are torn between the need to expand TIL quickly to maintain a ?young? phenotype that avoids exhaustion and the induction of tolerance, and the need to enrich for tumor- reactive TIL. We propose a novel technology called autohistomagnetic isolation (AHMI) that uses patient tumor-derived antigen presentation complex to select for reactive TIL within the first 48 hours after biopsy. The basic principle of this new methodology uses conformation-dependent antibodies to immunoprecipitate heterotrimers of HLA or H2-Kb, ?-2 microglobulin, and cognate peptide onto magnetic beads. The resulting construct can then isolate tumor-reactive TIL from non-reactive TIL using any magnetic separation platform among positively selected bulk TIL cultures. This will greatly reduce the time needed to generate TIL infusion products, will result in a more potent and persistent TIL phenotype, reduce cost, and increase anti-tumor reactivity. We propose to fully optimize AHMI and interrogate the subsequent TIL infusion product in a pre- clinical murine model of TIL therapy, and in samples from patients with metastatic melanoma. The completion of the studies proposed here will produce a fully functional prototype for mouse and human AHMI and lay the groundwork necessary for clinical trials.